Key facts
- Google DeepMind CEO Demis Hassabis believes humanity has a limited time to prepare for the arrival of AGI.
- Hassabis proposed the creation of a US-led coalition to review and ensure the safety of advanced AI models.
- He suggested an industry-funded standards body, similar to FINRA, to oversee AI model testing.
- Hassabis indicated that AGI could be here in a few short years, emphasizing the urgency of the situation.
Demis Hassabis, CEO of Google DeepMind, has issued a stark warning regarding the imminent arrival of artificial general intelligence (AGI), stating that humanity possesses a "precious window" to ensure its safety. In a recent essay, Hassabis proposed the establishment of a US-led coalition to review the most advanced AI models before their public release, drawing parallels to the Financial Industry Regulatory Authority (FINRA) as a model for an industry-funded standards body.
Hassabis indicated that AGI could be just a few short years away, a timeline he has previously suggested could be as early as 2030. He stressed the need for a "dynamic, adaptable, and rigorous" approach to testing frontier AI capabilities, arguing that the US is well-positioned to lead the development of such a framework. This proposal follows recent actions by the Trump administration to limit the release of AI models by companies like OpenAI and Anthropic.
He also highlighted that the profound implications of AGI, touching upon fundamental questions about human values, meaning, and purpose, cannot be left solely to technologists. Hassabis called for a societal-wide effort to define this new era, emphasizing that "what we collectively do now will determine how the next phase of civilisation unfolds."